AI 中文总结
TI-StegoAlign是一种仅更新LoRA参数的通道导向后训练框架，通过BCSO和CCPO优化，在分词不一致场景下实现100%接收方比特准确率，同时降低归一化困惑度偏差并提升抗隐写分析性能。
AI 中文摘要
生成式文本隐写使大语言模型（LLM）智能体能够通过与任务相关的消息交换秘密信息。然而，大多数方法在发送方侧的分词上评估恢复效果，而接收方仅能观察到表面文本。去分词和接收方侧的重新分词会改变分词边界、使编码状态不同步，导致此类评估高估接收方侧的恢复效果。现有补救措施依赖推理时的过滤或验证，仅校正单个输出，未使生成策略适配接收方侧通道。为解决这些局限，我们提出TI-StegoAlign，一种通道导向后训练框架。位一致性监督目标（BCSO）在已实现的发送方侧嵌入位置处扩大局部编码裕度；通道条件偏好优化（CCPO）随后利用接收方真实恢复效果、文本质量及抗隐写分析反馈对齐完整隐写文本。TI-StegoAlign仅更新LoRA参数，通信期间无需特定分词校正。实验结果显示，其接收方比特准确率达100%；与最强基线相比，TI-StegoAlign使归一化困惑度偏差降低21.6%，抗隐写分析性能相对提升6.3%。
英文摘要
Generative text steganography enables LLM agents to exchange secret information through task-relevant messages. Yet most methods evaluate recovery on sender-side tokens, whereas the receiver observes only surface text. Detokenization and receiver-side retokenization can alter token boundaries, desynchronize coding states, and cause such evaluation to overestimate receiver-side recovery. Existing remedies rely on inference-time filtering or verification, correcting individual outputs without adapting the generation policy to the receiver-side channel. To address these limitations, we propose TI-StegoAlign, a channel-guided post-training framework. The Bit-Consistent Supervised Objective (BCSO) enlarges local coding margins at realized sender-side embedding positions. Channel-Conditioned Preference Optimization (CCPO) then aligns complete stegotexts using receiver-realistic recovery, text quality, and anti-steganalysis feedback. TI-StegoAlign updates only LoRA parameters and requires no tokenization-specific correction during communication. Experimental results show 100% receiver bit accuracy. Compared with the strongest baselines, TI-StegoAlign achieves a 21.6% reduction in normalized perplexity deviation and a 6.3% relative improvement in anti-steganalysis performance.
